Soft latch push button switch - works with some loads but not with others

Thread Starter

Zurn

Joined Mar 4, 2019
117
I'm wondering if someone can point me in the right direction towards understanding:

I've assembled this soft latch push button switch (from the EEV blog) :


It works with an LED, or a small current consuming load. However, I'm trying to use it to power on/off a wifi router (5V, about 300 mA) and it won't latch in the OFF position.

I'm currently researching this on my own but the sooner I figure it out the better, so if anyone has any clues it'd be greatly appreciated.
 

AlbertHall

Joined Jun 4, 2014
12,345
It may be that he router has a capacitor on the input and this holds up the output voltage so it switches on again.
You could confirm this by connecting a capacitor, say 100uF, to the output with the LED and see if you get the same problem.
To get around this you could connect the load via a diode. You will lose some voltage this way, about 0.7V with a standard silicon diode or 0.4V with a schottky diode.
 

Thread Starter

Zurn

Joined Mar 4, 2019
117
Ya, the router definitely has a cap in it for power bumps. Adding the 100uF cap to the LED circuit produces the exact same effect.

I'll try the diode circuit route, though it will involve switching my voltage regulator to make up for the loss (I'll have to pump in 5.7V). I do need 5V exactly, so in the mean time if there's an alternative I'd love to hear it. I'm not opposed to building an entirely different latch circuit.
 

Thread Starter

Zurn

Joined Mar 4, 2019
117
Having tried the diode... it works... but it's a little slippery. I have to hold the button down for a few seconds, and during that time there's some serious oscillation that I'd rather not have on my load.

Alternatively, a different latch circuit powered only from the IN side of the FET, would avoid the need for that diode.
If you can direct me towards one it'd be greatly appreciated. Still searching/learning on the web in the mean time...
 

AlbertHall

Joined Jun 4, 2014
12,345
Having tried the diode... it works... but it's a little slippery. I have to hold the button down for a few seconds, and during that time there's some serious oscillation that I'd rather not have on my load.
I think that is because then it has no load once the output starts to fall and the diode is reverse biased so the output falls only slowly.
You could add the LED before the diode and you get a 'free' on/off indicator.
 

Thread Starter

Zurn

Joined Mar 4, 2019
117
I think that is because then it has no load once the output starts to fall and the diode is reverse biased so the output falls only slowly.
You could add the LED before the diode and you get a 'free' on/off indicator.
ya, moving the diode to the ground side of the load does the trick.
 

Thread Starter

Zurn

Joined Mar 4, 2019
117
Both look great... I've already wired it up w/the new adjustable voltage regulator to compensate for the diode loss. HOWEVER, I'll definitely checking these out in the near future.

OKAY if anyone cares to explain what this diode is actually doing I'd love to know... I don't quite grasp the theory of how it's allowing the switch to latch.
 

AlbertHall

Joined Jun 4, 2014
12,345
When you press to switch it off, the diode prevents the voltage from the capacitor in the router feeding back into the switch circuit and so allows it to switch off.
 

crutschow

Joined Mar 14, 2008
34,283
Alec_t's referenced circuit in post #9 has the advantage of the toggle being independent of the load characteristics.

Below is the LTspice simulation of a similar circuit:

upload_2019-9-12_13-34-20.png
 
Last edited:

Thread Starter

Zurn

Joined Mar 4, 2019
117
For the record, I ended up using the circuit referenced by Alec_t in post #9 and it's working great so far.
 

jpanhalt

Joined Jan 18, 2008
11,087
Sorry, I caught up to this thread a bit late. But here is a summary of my files on the subject:

First, here is a link to Dave's design:
Be sure to read the suggested improvements.

Second, here is a nice review of various methods: http://www.mosaic-industries.com/em...n-switch-turn-on/latching-toggle-power-switch

Finally, I have decided on the MAX16054 for a design (https://datasheets.maximintegrated.com/en/ds/MAX16054.pdf ). It incorporates debouncing on the input. Dave's does not. Debouncing of a non-powered system was important to me. The LTC2955 and maybe a lot of others do that too.
 
Top